Peter Danielson

This book explores the role of artificial intelligence in the development of a claim that morality is person-made and rational. Professor Danielson builds moral robots that do better than amoral competitors in a tournament of games like the Prisoners Dilemma and Chicken. The book thus engages in current controversies over the adequacy of the received theory of rational choice. It sides with Gauthier and McClennan, who extend the devices of rational choice to include moral constraint. Artificial Morality goes further, by promoting communication, testing and copying of principles and by stressing empirical tests.

Is game theory good for us? This may seem an odd question. In the strict sense, game theory—the axiomatic account of interaction between rational agents—is as morally neutral as arithmetic. But the popularization of game theory as a way of thinking about social interaction is far from neutral. Consider the contrast between characterizing bargaining over distribution as a “zero-sum society” and focussing on “win-win” cooperative solutions. These reflections bring us to the book under review, Prisoner's Dilemma, a popular introduction to game theory and its relation to ethics by the respected science writer William Poundstone. The book begins with a moral dilemma and ends by discussing the evolution of co-operation. Poundstone emphasizes—correctly, to my mind—the ethical potential of game theory. He concludes his first chapter with this striking claim: “Today's practitioners of game theory are attempting to forge a kind of ethical progress. Is there any way to promote the common good in a prisoner's dilemma? The attempt to answer this question is one of the great intellectual adventures of our time”.

Emerging technologies like robotics for war and peace stress our moral norms and generate much public interest and controversy. We use this interest to attract participants to an innovative on-line survey platform, designed for experimenting with public engagement in the ethics of technology. In particular, the N-Reasons platform addresses several issues in democratic ethics: the cost of public participation, the methodological issue of feasible reflective ethical equilibrium (how can individuals in a large group, take into account the ethical views of all others?), and the reliability of public participation processes. We sketch the motivation and design of the N-Reasons platform, stressing the need for a practical (fast, low-cost) instrument that makes equilibrium feasible. We focus on the Robot Ethics Survey that featured a set of nine ethical challenges raised by robotics for war and peace. Over 400 people in five disjoint groups participated in this on-line survey experiment. We analyze the results, both quantitatively and qualitatively, the participants’ decisions taken and the reasons supporting these decisions. Both decisions and reasons strongly distinguished lethal military robotics from peace-related robotics. Methodologically, both decisions and reasons over five distinct groups were remarkably consistent.

Background: Recent studies on public attitudes toward life extension technologies show a mix of ambivalence toward and support for extending the human lifespan. Attitudes toward genetic modification of organisms and technological enhancements may be used to categorize individuals according to political or ideological orientation such as technoprogressive or conservative and it could be easy to assume that these categories are related to more general categorizations related to culture, e.g. between Traditional and Secular-rational values in the World Values Survey. This paper discusses how attitudes toward aspects of radical life extension may be related to cultural values as revealed in an online deliberative survey among university students conducted between January 2012 to January 2013. Survey results suggest that attitudes toward radical life extension tend to be mixed among groups categorized as Traditional, Secular-rational, Survivalist, and Self-expressionist. The study explored the relation between responses of 326 university students to 5 key questions on radical life extension and the cultural values they tend to favor as indicated by their response to 20 statements from the World Values Survey.Design and Method: The survey consisted of 3 stages: an online pre-discussion survey, face-to-face discussion, and post-discussion survey. After completing the 5 main survey questions in stage 1, participants were presented two additional questionnaires: one on cultural attitudes using 20 statements from the 2004-2008 World Values Survey and another on health attitudes with 12 statements from Dutta-Bergman’s 2004 study. In stage 2, participants were engaged in a face-to-face discussion in class focusing on their responses to the five key questions. After the discussion, they were invited to reconsider the choices and reasons they posted in stage 1 in the light of the face-to-face class discussion in stage 2.Results: Responses to the five survey questions showed that there tended to be more individuals across groups who disagreed with adopting technologies that radically extend the human lifespan beyond the current limit of 120 years. Attitudes toward radical life extension did not correspond to cultural attitudes indicated by responses to the WVS questions. The proportion of agreement/disagreement to statements presented in each of the five questions varied across cultural groups and there tended to be more individuals who disagreed with radical life extension in all groups. Changes in responses after the discussion stage were not significant and most respondents maintained their prior views.Discussion: Cultural attitudes associated with familiar technologies may not correspond with attitudes toward newer technologies since beliefs and values may need to be adapted to new imagined situations that the new technologies elicit. Moral understandings associated with familiar technological habits and beliefs are not necessarily carried over to new technologies.

Rationality and evolution are apparently quite different, applying, respectively, to the acts of complex, well-informed individuals and to populations of what may be mindlessly simple entities. So it is remarkable that evolutionary game theory shows the theory of rational agents and that of populations of replicating strategies to be isomorphic. Danielson illustrates its main concepts—evolutionarily stable strategies and replicator dynamics—with simple models that apply to biological and social interactions; and he distinguishes biological, economic, and generalist ways of interpreting the theory. Against the background of isomorphism, he considers three ways in which evolution and rationality differ and how two-level models may combine them. He concludes with a survey of the normative significance of the unification of rationality and evolutionary game theory and some speculation about the evolution of human rationality.

N-Reasons is an experimental Internet survey platform designed to enhance public participation in applied ethics and policy. N-Reasons encourages individuals to generate reasons to support their judgments, and groups to converge on a common set of reasons pro and con various issues. In the Robot Ethics Survey some of the reasons contributed surprising judgments about autonomous machines. Presented with a version of the trolley problem with an autonomous train as the agent, participants gave unexpected answers, revealing high expectations for the autonomous machine and shifting blame from the automated device to the humans in the scenario. Further experiments with a standard pair of human-only trolley problems refine these results. While showing the high expectations even when no autonomous machine is involved, human bystanders are only blamed in the machine case. A third experiment explicitly aimed at responsibility for driverless cars confirms our findings about shifting blame in the case of autonomous machine agents. We conclude methodologically that both results point to the power of an experimental survey based approach to public participation to explore surprising assumptions and judgments in applied ethics. However, both results also support using caution when interpreting survey results in ethics, demonstrating the importance of qualitative data to provide further context for evaluating judgments revealed by surveys. On the ethics side, the result about shifting blame to humans interacting with autonomous machines suggests caution about the unintended consequences of intuitive principles requiring human responsibility.